The capillary endothelial enzyme, xanthine oxidase/dehydrogenase, which mediates oxidation of hypoxathine to xanthine, and xanthine to uric acid, was investigated in the perfused beating heart. Intracoronary bolus injections were made of [14C]-hypoxanthine (together with reference tracers [131I]-albumin and [3H]-L-glucose), while measuring venous concentrations of the injected tracers and [14C]-xanthine and [14C]-uric acid, formed in endothelial cells. Sufficient unlabeled carrier hypoxanthine was co-injected that peak concentrations swept through the likely concentration range of the enzyme Km. The dilution curves showed strikingly low levels of coronary efflux of [14C]-xanthine, suggesting very low endothelial membrane permeability. However, the possibility of low permeability had to be discarded as a result of additional experiments in which [14C]-xanthine was injected in the tracer bolus instead of hypoxanthine, and it was observed that the peak extraction of xanthine was nearly as high as that of hypoxanthine (0.5). Fitting the dilution curves using a multiple species nonlinear model of capillary-tissue transport, binding and reaction indicated that prolonged retention of xanthine following hypoxanthine injection was likely due to slow dissociation of the enzyme-xanthine product complex. This appears to favor subsequent reaction to uric acid over release of free xanthine. Consistent with this interpretation was the finding that when [14C]-xanthine was injected in the bolus, there was less formation of end product [14C]-uric acid, than when [14C]-hypoxanthine was injected. Therefore, it seems that the high affinity binding of the enzyme to xanthine serves to insure that most of the hypoxanthine oxidized in the first step of the reaction is not released from the enzyme until it is further oxidized to uric acid in the second step.